|Brain image segmentation method with hybrid model based on FESS|
|*Lian, Yuanfeng; Zhao, Yan; He, Huiguang; Chen, Xuejiao
|Source Publication||Chinese Journal of Scientific Instrument
|Abstract||This paper presents a new brain image segmentation method based on free energy score space (FESS), which combines the advantages of both generative model and discriminative model. Through conditional random field, the generative model fuses the grey scale information, shape information and spatial neighborhood relation of the volume pixels; and the appearance feature description of subcortical structure is realized. Based on the above, the generative model is used to map the training samples to FESS feature space. Furthermore; in the discriminative model, the LS-SVM classifier is used and the data field is applied in the training process of the mixed segmentation model, which reduces the performance fluctuation of the discriminative model induced by the imbalanced training data sets and effectively improves the generalization capability. Experimental results show that the proposed model possesses better segmentation quality and performance compared with several other state-of-the-art brain image segmentation approaches.|
|Keyword||Conditional Random Field
Least Squares Support Vector Machines
*Lian, Yuanfeng,Zhao, Yan,He, Huiguang,et al. Brain image segmentation method with hybrid model based on FESS[J]. Chinese Journal of Scientific Instrument,2013,34(6):1226.
*Lian, Yuanfeng,Zhao, Yan,He, Huiguang,&Chen, Xuejiao.(2013).Brain image segmentation method with hybrid model based on FESS.Chinese Journal of Scientific Instrument,34(6),1226.
*Lian, Yuanfeng,et al."Brain image segmentation method with hybrid model based on FESS".Chinese Journal of Scientific Instrument 34.6(2013):1226.
|Files in This Item:||
||There are no files associated with this item.
Items in the repository are protected by copyright, with all rights reserved, unless otherwise indicated.